Pascal’s recent thoughts on 3D (#1181 on 3D and pop) are a tough act to follow! As announced, here is a second instalment with a slight change of subject—or at least of approach. In this post I’ll write in my own allegedly “academic” way, seeking to separate out and analyze a few factors, and hope that our two different approaches will both be useful. Our images may be different, but I believe our ideas are in harmony.
How can we see distance in the flat surface of a photograph? Let’s take a step back and ask: how can we see distance with our eyes? After all, our retinas are flat surfaces (any bumps would occasion a visit to the ophthalmologist), yet most of us see a wonderfully 3-D world. Our complex visual apparatus has a number of tricks to get from there to here: it utilizes different kinds of receptors, parallax, focus, eyeball saccades, proprioceptive data, and much more. In addition, the brain has some amazing algorithms to sort through the huge amount of data. In fact, seeing distance is greatly over-determined—as well it might be considering its importance. After all, it can be a matter of life and death to accurately determine the distance of a predator. We trust our eyes implicitly, although maybe we shouldn’t be overly confident that visual appearance is reality.
Photographs are two-dimensional like retinas, yet they usually portray a three-dimensional world (abstractions excepted). How is this possible? Well, it’s easier in stereoscopy, where two different lenses provide sufficient parallax to fool our eyes when properly viewed. It’s also easier in video: where time is added there is movement, either in the scene or by the camera (including change of focus), again fooling the eye.
But how can we suggest distance or depth in a still photograph? Surprisingly, there are quite a few cues, most of which can work together to give a strong and natural impression of a third dimension. I’ll present a list which complements Pascal’s. Though I have tried to be comprehensive, my list is surely incomplete, and you are welcome to add more items, or perhaps to slice the cheese differently.
(i) It’s hard to find images that illustrate only one of the cues, so my examples are ones in which the feature in question is prominent, not exclusive.
(ii) These different cues are for analytical purposes only. Photographers in the field won’t consult them, but they may upon reflection realize they’ve been using some of them all along, and they can think about looking for others next time.
(iii) Consider my comments as tentative generalizations (a phrase that many people don’t seem to realize is redundant), open to counter-examples but still providing useful summaries.
(iv) All my images were taken with one camera and one lens, a Sony A7III with a walkabout 24-105 mm zoom, in contrast to Pascal’s arsenal.
More distant objects tend to be less intense and on the blue-violet end of the spectrum. Sometimes the subject of interest is in the distance, like sunrises and sunsets, or mountains far away lit by slanting sunlight, and then we will try to make them as light and clear as we can. But mostly we are interested in something closer, and color is sometimes useful for indicating this.
Here the distant mountains fade into blue, obviously farther than colorful foliage nearby. This is a view from the Blue Ridge Parkway looking west, late afternoon in mid-November.
Overlap or occlusion
Nearer things block out further things on the same line of sight. This is fundamental to the world we construct from visual data.
Forests are paradigm examples of occlusion, and trees that block out others are clearly nearer. This is the local forest in mid-February, with trees advancing down the hillside. The distant hills are also bluer as well as being occluded by the trees.
Larger things seem nearer than smaller ones.
These tree trunks, taken in early March, rely on occlusion, texture and focus as well as size, but size does duty for nearness as well (I think the diameter of the two trees is actually about the same).
There are three kinds of perspective at play in photographs (actually four, if you count other kinds of leading lines), and they may be variously combined.
Aerial or atmospheric perspective: Distant objects are normally less distinct, blurred by haze. This blurring is literally in the air, and not in the camera like bokeh.
This image of House Mountain was taken from North Mountain last December; the Blue Ridge Mountains in the background are blurred and less distinct. Color and occlusion help as well.
Renaissance painters famously exploited this form of perspective, where straight lines converge at the horizon, to give their images (the illusion of) depth.
This footbridge was installed over the South River late last year. The girders are maybe a little too obvious in their convergence, but they leave no doubt about what is near and what is far.
Sometimes the lines aren’t straight Euclidean ones but curved; this can be obtained with fisheye lenses or extreme panoramas. Some dislike the effect, but it can be interesting, and certainly conveys depth.
This Rockbridge County road in January does indeed curve around before me, but the effect is enhanced by the panorama. The mountains in the background covey depth as well, since they are familiar objects that are less distinct than the field, although they are snowy white, not distant blue.
These can be straight or curved lines, as before, but they can also zigzag their way into the distance. They lead the eye (or the eye perceives them as leading) into the distance.
This street runs in front of our cottage, but it doesn’t run straight into the distance, it curves. I desaturated the image because then you won’t focus on the old dirty snow. There are quite a few distance markers in this image.
Shades and shadows
Lighter tones usually, though not invariably, suggests nearer subjects, darker tones more distant ones. Sunrises and sunsets are prime exceptions.
This is not a particularly good image, but it illustrates the point: the crocuses are lighter than the distant house.
The hills are a mile or so distant, but there’s a lot going on in the sky, which is even more distant: clouds, contrails and the afterglow of sunset.
Elevation above the horizon suggests distance, while closer things are usually beneath the horizon.
These hills (upgraded to mountains in the local lexicon) undulate into the distance, and the higher they are placed in the photograph the more distant they are.
This photograph (of a photographer taking a photograph) of the California coast at twilight has several distance cues: the prominent leading line of course, the color gradient, but also the height of the islands in the photograph. In the dim light, I had to bump the ISO to 25,600; it’s a wonder there was a usable image at all.
More finely detailed objects appear nearer than more distant ones. After all, we’re accustomed to the acuity of our vision falling off with distance.
This nearby forest path has a leading line and occlusion of tree trunks, but also the nearer trunks are more finely detailed (especially because they are more in focus).
Objects we can readily identify may appear closer than ones that are strange to us.
These vultures stretching their wings to dry after a snowstorm are familiar enough that we can place their distance reasonably well, and they may even appear closer than the fence, which they are not.
Connectedly, familiar objects can help us size up unfamiliar objects and hence place them at the proper distance.
We (think we) know the size of cows, though they are really quite large animals, and their relative size in this image is one of many distance cues. Others include the leading line of the trail, the dark hills, and the sky, a form of linear perspective.
I saved this one for last, because it is over-hyped by fast-lens fanboys and doesn’t always convey distance. There are degrees of defocus, of course, and while a touch of bokeh can suggest depth, extreme bokeh just yields a fuzzy background, good for isolating portraits perhaps but not very helpful for depth. Bokeh works best with other cues such as shades, textures and occlusion.
These three trees are in differing focus, with more detailed texture in the closest trunk. The curved twig is a leading line from tree #1 to tree #2.
I close with a few “flat” images for contrast. They illustrate the subject by their relative paucity of distance cues.
How far away is this rock face? Actually it’s only about 10 feet distant, but it could well have been further. I wanted a drab palette, as well-focused as I could manage under the conditions (24 mm, f/10). The trees do indicate a receding hill, but overall it still seems quite flat to me.
How far away are these (mostly sycamore) trees? Actually they are about a quarter mile distant, across an unseen pasture and river. That’s a four-lane highway running through the trees. Everything is flattened in telephoto land.
I deliberately inverted this rock face image to remove some light and dark distance cues, with only the vines superimposed on the cliff providing a smidgeon of depth. It’s verging on an abstraction.
Never miss a post
Like what you are reading? Subscribe below and receive all posts in your inbox as they are published. Join the conversation with thousands of other creative photographers.
Please log in again. The login page will open in a new tab. After logging in you can close it and return to this page.